Enhanced PV Power Prediction Considering PM10 Parameter by Hybrid JAYA-ANN Model

Abstract

The demand for electrical energy is continuously increasing in these days, particularly due to advancements in the industrial sector. This surge in demand has underscored the importance of seeking alternative energy sources, with solar energy emerging as a standout option due to its low investment costs and environmental friendliness. However, the variability in photovoltaic power production, influenced by meteorological data, necessitates accurate prediction methods. To enhance the precision of these predictions, incorporating new parameters alongside existing meteorological data is advantageous. In this regard, this study explores the impact of the particulate matter (PM10) parameter on photovoltaic power prediction using artificial neural network (ANN) model and JAYA-ANN. Comparing the prediction results based on root mean squared and mean absolute percentage errors reveals that the hybrid JAYA-ANN model consistently outperforms the ANN and persistence models. Notably, the PM10 parameter proves to be a significant input in forecasting daily photovoltaic power.

Keywords:

• Photovoltaic power production
• metaheuristic optimization
• artificial neural network
• prediction
• comparison

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Erdal Irmak

Erdal Irmak (Senior Member, IEEE) earned his BSc degree in Electrical Education from the Faculty of Technical Education at Gazi University, Türkiye, in 1997. Subsequently, he pursued his MSc and PhD degrees in the same field at the Gazi University Graduate School of Natural and Applied Sciences, in 2001 and 2007, respectively. His expertise covers power system operation and control, voltage and frequency stability, the integration of renewables into power systems, the design of virtual and remote labs, and the cybersecurity of critical infrastructure. He has authored or co-authored over 120 research papers, most of which are indexed in WoS. Dr. Irmak has led and contributed to various projects, spanning from solar cell technology to microgrid development and real-time pricing optimization in smart grids. He has played key roles in IEEE-supported conferences including GPECOM, ICRERA, and ICSMARTGRID. He serves as the Editor of EPCS and as Associate Editor in prestigious journals such as IJRER and GUJSc. Currently, he is full-time Professor in Electrical and Electronics Engineering Department of Gazi University.

Mehmet Yeşilbudak

Mehmet Yeşilbudak received his B.Sc. and Ph.D. degrees from Gazi University, Ankara, Turkey, in 2008 and 2014, respectively. He currently works as an Associate Professor at the Department of Electrical and Electronics Engineering, Nevsehir Haci Bektas Veli University, Nevsehir, Turkey. His current research interests include renewable energy sources, wind and solar energy systems, artificial intelligence, data mining and metaheuristic optimization.

Oğuz Taşdemir

Oğuz Taşdemir received his bachelor’s degree from Kırıkkale University, Faculty of Engineering, Department of Electrical and Electronics Engineering in 2010 and his master’s degree from Nevsehir Haci Bektas Veli University, Institute of Science, Department of Electrical and Electronics Engineering in 2020. He started his PhD education in 2020 at Gazi University, Graduate School of Natural and Applied Sciences, Department of Electrical and Electronics Engineering in Ankara, Turkey. His main research interests include power quality problems, artificial neural networks, photovoltaic energy systems and metaheuristic algorithms. He is currently working as a lecturer at Kırsehir Ahi Evran University Kaman Vocational High School, Department of Electricity and Energy.